Design of an Expert System for Distribution Planning System using Soft Computing Techniques

Design of an Expert System for Distribution Planning System using Soft Computing Techniques

Shabbir Uddin (Department of Electrical and Electronics Engineering, Sikkim Manipal Institute of Technology, Sikkim, India), Amitava Ray (Department of Mechanical Engineering, National Institute of Technology, Silchar, India), Karma Sonam Sherpa (Department of Electrical and Electronics Engineering, Sikkim Manipal Institute of Technology, Sikkim, India) and Sandeep Chakravorty (Department of Electrical Engineering, Baddi University of Emerging Science and Technology, Baddi, India)
Copyright: © 2016 |Pages: 19
DOI: 10.4018/IJEOE.2016040103
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

An expert system for distribution planning is proposed in this paper. Choosing a best location of a distribution substation and grouping the various load points to be fed from a particular distribution substation has always been a concern to the distribution planners. Here in this paper the authors present a hybridization of K-means clustering method with fuzzy context aware decision algorithm for choosing the optimum location of distribution substation and its feeder layout. K means clustering has been applied to various loads which are at different location to form a cluster with load points in closer proximity so that a substation could be placed for each cluster for the distribution of power. Fuzzy Context Aware Decision Algorithm based on the Analytical Hierarchy process (AHP) is then applied on each cluster to decide on the feeder layout connecting the load points in each cluster. The feeder layout is based on the various reliability factors and thus the result obtained will lead to optimum feeder path and will hence lower long range distribution expenses.
Article Preview

1. Introduction

The problem of distribution planning involves optimizing the selection of locations and sizes of new substation and feeder to meet the requirement of load demands under the constraints of substation capacity, voltage drop, feeder thermal capacity, radial network configuration and reliability. The planning, design and operation of power distribution system require continuous and comprehensive analysis in order to evaluate the system performance and to determine the effectiveness of alternative plans for system expansion. These studies play a vital role in providing a high standard of power system reliability, security and quality and in ensuring the maximum utilization of optimal investment. For several years, deterministic methods based on engineering judgment and experiences, and also on worst case conditions, have been used successively in power system planning. However we see that number of factors such as losses in the system, cost, reliability etc plays and effect the system planning, in other words we can say that multiple criteria needs to be considered before we converge to an optimum placement of substation and conductor layout.

In general, the decisions in the planning of power distribution system include:

Optimal location of substations, optimal allocation of load and optimal allocation of substation capacity.

The available literature consists of work of only few researchers on the field of distribution planning. Most of them are based on mathematical programming such as transshipment algorithms, transportation (Crawford and Holt, 1975; El-Kady, 1984), mixed integer programming (Gonen and Ramirez-Roasado, 1987), dynamic programming (Partanen, 1990) etc. K. K. Li and T. S. Chung (2004) proposed genetic algorithm to find the optimum location of substation to meet the load demands of 13 load points whose coordinates and MVA demands are given. Same kind of work has been carried out by Belgin Turkay and Taylan Artac (2005), the work by J. F. Gomez et al. (2004) is also on the same grounds. In all the above cases planning of laying the feeders or distribution planning has been done either by man machine interface or heuristic algorithm.

The techniques for the solution of the planning problem of primary distribution circuits can be found in Khator and Leung (1997) and Bernal-Agustin (1998). Techniques such as Heuristic search methods have been developed (Boardman and Meekiff, 1985; Nara et al., 1992) showing faster performance than the conventional optimization techniques but with some limitations in the goodness of the solutions to the problem that are obtained. In Bernal-Agustin (1998) and Carvalho and Ferreira (1998) the latent of the GA’s is shown in comparison with classical optimization techniques to solve the planning problem in a very complete and detailed formulation considering the nonlinearity of the cost function, the limits of the voltage magnitudes and a term in the objective function to take into account the reliability of the system, reporting significant improvements in the solution times. An integer variable coding scheme was used to facilitate the consideration of different conductor sizes and substation sizes also new genetic operators were proposed to improve the performance of the algorithm. In Dorigo and Gambardella (1997) an approach is expanded to consider the multiple development stages as well as multiple objectives. An evolutionary approach (Diaz Dorado et al., 2002) is applied to the design of a medium voltage network using a detailed model of the network.

Complete Article List

Search this Journal:
Reset
Open Access Articles
Volume 6: 4 Issues (2017)
Volume 5: 4 Issues (2016)
Volume 4: 4 Issues (2015)
Volume 3: 4 Issues (2014)
Volume 2: 4 Issues (2013)
Volume 1: 4 Issues (2012)
View Complete Journal Contents Listing